Compare AI models

Specifications
Parameters
—0.6B/1.7B/4B/8B/14B/30B/32B/235B
Context window
1M—
API pricing
Cheapest input
$0.10Azure$0.064DeepInfra
Cheapest output
$0.40Azure$0.192DeepInfra
$40.00Azure$30.00DeepInfra · fp8
Benchmarks
IFEval
74.5%83.4%
MATH-500
62.3%98%
AIME 2024
29.4%85.7%
GPQA Diamond
50.3%71.1%
Overview
CompanyOpenAIQwen
Release dateApr 14 2025Apr 29 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen3 leads GPT-4.1 nano on 4 of the 4 benchmarks they both report (IFEval, MATH-500, AIME 2024, GPQA Diamond). GPT-4.1 nano shipped 15 days before Qwen3, so benchmark comparisons should account for the intervening progress.

GPT-4.1 nano is closed, while Qwen3 is open weight.

On IFEval, Qwen3 leads at 83.4% vs GPT-4.1 nano at 74.5%. On MATH-500, Qwen3 leads at 98% vs GPT-4.1 nano at 62.3%. On AIME 2024, Qwen3 leads at 85.7% vs GPT-4.1 nano at 29.4%. On GPQA Diamond, Qwen3 leads at 71.1% vs GPT-4.1 nano at 50.3%.